Disentangling shallow‐water bulk carbonate carbon isotope archives with evidence for multi‐stage diagenesis: An in‐depth component‐specific petrographic and geochemical study from Oman (mid‐Cretaceous)

Disentangling shallow‐water bulk carbonate carbon isotope archives into primary and diagenetic components is a notoriously difficult task and even diagenetically screened records often provide chemostratigraphic patterns that significantly differ from global signals. This is mainly caused by the polygenetic nature of shallow‐water carbonate substrates, local carbon cycle processes causing considerable neritic–pelagic isotope gradients and the presence of hiatal surfaces resulting in extremely low carbonate preservation rates. Provided here is an in‐depth petrographic and geochemical evaluation of different carbonate phases of a mid‐Cretaceous (Barremian–Aptian) shallow‐water limestone succession (Jabal Madar section) deposited on the tropical Arabian carbonate platform in Oman. The superposition of stable isotope signatures of identified carbonate phases causes a complex and often noisy bulk carbon isotope pattern. Blocky sparite cements filling intergranular pores and bioclastic voids evidence intermediate to (arguably) deep burial diagenetic conditions during their formation, owing to different timing or differential faulting promoting the circulation of fluids from variable sources. In contrast, sparite cements filling sub‐vertical veins reveal a rock‐buffered diagenetic fluid composition with an intriguing moderate enrichment in 13C, probably due to fractionation during pressure release in the context of the Miocene exhumation of the carbonate platform under study. The presence of abundant, replacive dedolomite in mud‐supported limestone samples forced negative carbon and oxygen isotope changes that are either associated with the thermal breakdown of organic matter in the deep burial realm or the expulsion of buried meteoric water in the intermediate burial realm. Notwithstanding the documented stratigraphically variable and often facies‐related impact of different diagenetic fluids on the bulk‐rock stable isotope signature, the identification of diagenetic end‐members defined δ13C and δ18O threshold values that allowed the most reliable ‘primary’ bulk carbon isotope signatures to be extracted. Most importantly, this approach exemplifies how to place regional shallow‐water stable isotope patterns with evidence for a complex multi‐stage diagenetic history into a supraregional or even global context.


INTRODUCTION
Shallow-water carbon isotope variations recorded by bulk carbonate material have been proven to provide welldefined tie points for the correlation of deep-time carbonate platform sections and has further been used to stratigraphically link major changes in biosedimentation (biotic turnovers, carbonate platform demise and drowning) to climate and palaeoenvironmental changes associated with major carbon cycle perturbations (F€ ollmi et al., 1994Ferreri et al., 1997;Gr€ otsch et al., 1998;Wissler et al., 2002;Immenhauser et al., 2005;Parente et al., 2007;Burla et al., 2008;Frijia & Parente, 2008;Elrick et al., 2009;Mill an et al., 2009Mill an et al., , 2011El-Sabbagh et al., 2011;Huck et al., 2011Huck et al., , 2013Huck et al., , 2014Di Lucia et al., 2012;Krencker et al., 2014;Bodin et al., 2015;Wohlwend et al., 2016). Unfortunately, the carbon isotope-based stratigraphic refinement of biostratigraphically poorly constrained shallow-water sections is often afflicted with large uncertainties unless additional stratigraphic methods such as strontium isotope stratigraphy (SIS) are applied (Huck et al., 2010(Huck et al., , 2011Horikx et al., 2014;Frijia et al., 2015;Huck & Heimhofer, 2015;Bover-Arnal et al., 2016). Without a proper SIS framework, shallow-water carbon isotope stratigraphy is solely based on the correlation of similar fluctuations ('wiggle matching') within a dating uncertainty . This is particularly problematic, as the shallow-water carbon isotope pool of dissolved inorganic carbon (DIC) is affected to varying degrees by (i) local carbon cycle processes, (ii) syndepositional diagenesis and (ii) fractionation effects associated with varying contributions of aragonite, lowmagnesium and high-magnesium calcite (Immenhauser et al., 2002(Immenhauser et al., , 2008. Moreover, the shape of shallow-water carbon isotope curves is strongly modulated by sea-level fluctuations, as subaerial exposure and re-flooding of the platform causes phases of non-deposition, erosion, reworking and bypass of sediments. As a consequence, up to 90% of the carbon isotope signal may be lost in discontinuity surfaces or condensed in thin sedimentary layers (Strasser, 2015) and thus 'wiggle matching' between carbon isotope curves with locally different absolute values, amplitudes and gradients of excursions will allow multiple stratigraphic interpretations .
The current study aims at exemplifying how to extract stratigraphically meaningful global carbon cycle fluctuations out of shallow-water bulk carbonate carbon isotope archives that show evidence for a complex multi-stage diagenetic alteration. Therefore, a detailed sedimentological, petrographic and geochemical approach has been applied to a mid-Cretaceous southern Tethyan Arabian carbonate platform section (Jabal Madar) in Oman (Fig. 1A). The Jabal Madar section has been chosen as an ideal case study, as numerous biostratigraphic and sequence stratigraphic tie points are available Pittet et al., 2002;Schr€ oder et al., 2010), but the stratigraphic resolution, in particular with respect to the Barremian interval, is rather low. Owing to previous work (van Pittet et al., 2002), the studied limestone succession exposed at the Jabal Madar dome evidently suffered strong multi-stage diagenetic alteration in the eogenetic, mesogenetic and telogenetic realms (mud cracks, rhizoliths, stylolites, dolomite). Notably, the (compacted) carbonate 'preservation rate' sensu Strasser (2015) for the Barremian portion of the Jabal Madar section (calculated from Pittet et al., 2002) is low (0Á012 mm y À1 ), an observation that clearly highlights the dominance of phases of sediment starvation and/or removal.
The overall aim of this study is thus to disentangle the local bulk carbonate carbon and oxygen isotope pattern, that is, to evaluate the influence of (i) local carbon cycling processes and (ii) syn to post-depositional diagenesis on the d 13 C bulk record. Therefore, a component-specific comparison of carbon and oxygen isotope values is applied in order to define isotopic end-members representing specific diagenetic realms. In combination with an in-depth sedimentological, petrographic and elemental geochemistry inspection, a precise evaluation of the chemostratigraphic potential of the carbon isotope record is possible (Fig. 2). This will in turn allow a more objective correlation of the Jabal Madar d 13 C bulk record with local carbon isotope patterns representing different settings of the Arabian carbonate platform and thus, the integration of valuable stratigraphic information on a platform-wide scale. Based on this integrated chemostratigraphic, sequence stratigraphic and biostratigraphic framework, the Jabal Madar section can be placed against stratigraphically well-constrained Tethyan neritic and pelagic sections. The outcome of this study will help to better attribute Tethyan-wide major steps in carbonate platform evolution (e.g. orbitolinid mass occurrences) to their corresponding palaeoenvironmental and palaeoceanographic forcing mechanisms.

GEOLOGICAL SETTING
The Jabal Madar dome is located in the Adam Foothills of northern Oman, about 140 km south of Muscat and about 5 km east of 'Uyun (Fig. 1B). The base of the studied Jabal Madar section is situated in the eastern part of the dome (22°23 0 14Á79″N/58°10 0 03Á65″E). The section comprises Barremian to Aptian carbonate platform deposits belonging to the Lekhwair, Kharaib and Shu'aiba     Fig. 2. Flow chart illustrating the here applied diagenetic screening protocol that builds on a component-specific petrographic and geochemical approach. This approach allows shallow-water bulk carbonate carbon isotope records with evidence for a multi-stage diagenetic alteration to be evaluated.
A detailed sedimentological, sequence stratigraphic and cyclostratigraphic study of the Lekhwair, Kharaib and Shu'aiba formations exposed at the Jabal Madar dome is presented in Pittet et al. (2002).

FIELD APPROACHES AND LABORATORY METHODS
The working approach used for sedimentological characterization of the 74 m thick Jabal Madar section involved an outcrop-based carbonate facies description, supported by the petrographic analysis of 73 thin sections. Microfacies analysis followed the limestone classification scheme of Dunham (1962), including the modifications by Embry & Klovan (1971), and is based on a component analysis (including biostratigraphically meaningful microfossils) and on textural and diagenetic features. The percentage of individual carbonate phases including sparite cements and dedolomite rhombs present in selected thin sections (i.e. microphotographs) was estimated by applying the pixel counting method sensu Coimbra & Ol oriz (2012a). Samples for geochemical investigations were taken at a spacing of 0Á3 to 1Á5 m (mean: 0Á7 m). Higher sample densities were applied across facies boundaries and discontinuity surfaces.
Carbonate powders were extracted from carbonate slabs by means of a hand-held PROXXON IBS/E drill equipped with tungsten drill bits (maximum speed: 8000 rpm). In order to evaluate the intra-sample variability in the bulk carbonate carbon and oxygen isotope composition (Fig. 2), several subsamples were drilled from about 50% (n = 52) of all collected hand specimens (n = 115). In addition, sampling focused on the extraction of powders from the main carbonate phases identified at thin section scale. These include (i) diagenetic (dedolomitized limestone; sparite cement filling intergranular pore space, veinlets and voids) as well as (ii) near-primary carbonate phases (matrix micrite sensu stricto, low-Mg calcite bivalve shell fragments). In total, carbon and oxygen isotope analysis of 383 carbonate powder samples (bulk: n = 202; matrix micrite: n = 43; dedolomitized limestone: n = 36; sparite cement: n = 87; bivalves: n = 15) was performed at the isotope laboratories of the Institute of Geology at Leibniz University Hannover (LUH), Germany (Tables SI-1/2). Stable isotope analysis was conducted using a Thermo Fisher Scientific Gasbench II carbonate device connected to a Thermo Fisher Scientific Delta-V Advantage isotope ratio mass spectrometer. Aliquots of the samples (200 AE 30 lg) were treated with viscous water-free (98 g mol À1 ) orthophosphoric acid at 72°C to release CO 2 . In order to ensure that samples containing variable amounts of dolomite have proper equilibration times (>2 h) with the acid, the latter is injected manually before the start of the measurement. Repeated analyses of certified carbonate standards (NBS 19, show an external reproducibility of ≤0Á06& for d 13 C and 0Á08& for d 18 O. All isotope results are reported in per mil (&) relative to the Vienna-Pee Dee Formation belemnite (V-PDB) standard in the conventional manner. For chemostratigraphic correlation, a three-point moving average was calculated. Oxygen isotope ratios, plotted against carbon isotope ratios, are used as proxy for the impact of diagenetic alteration.
At the isotope laboratory of the Institute of Geology, Mineralogy and Geophysics at Ruhr-University Bochum (RUB), Germany, aliquots (1Á35 to 1Á65 mg) of 26 powdered subsamples (matrix micrite: n = 10; grainstone bulk carbonate: n = 5; sparite cement: n = 6; dedolomite-rich limestones: n = 3; bivalves: n = 3) were analysed for their major and trace elemental composition (calcium, magnesium, strontium, iron and manganese) using inductively coupled plasma-atomic emission spectrometry (ICP-AES). Strontium isotope ratios of selected subsamples (matrix micrite: n = 4; bivalve: 1) were measured at RUB by means of a thermal ionization mass spectrometer (Finnigan MAT 262) in dynamic mode. Corrections of measured strontium isotope ratios to a USGS EN-1 value of 0Á709175 were done following the procedure of Howarth & McArthur (1997). For more details on the analytical procedure please refer to Huck et al. (2011).
In addition, cathodoluminescence (CL) examination of 12 thin sections was carried out with a "hot cathode" CL microscope (type HC1-LM) at Potsdam University. The acceleration voltage of the electron beam is 14 kV and the beam current is set to a level gaining a current density of ca 9 lA mm À2 on the sample surface. Refer to Christ et al. (2012) for details on the analytical procedure.

LITHOSTRATIGRAPHY AND MICROFACIES OF THE JABAL MADAR SECTION
The studied 74 m thick Barremian portion of the shallow-water Jabal Madar section (total thickness: 111Á6 m; Pittet et al., 2002;Sattler et al., 2005) is dominated by restricted to open lagoonal fine-grained limestones with intercalations of coarse-grained partly cross-bedded, as well as slightly argillaceous orbitolinid-rich levels (Figs 3A and 4, Table 1). Dedolomitization is rather common and predominantly affects matrix micrites, burrow-infills and, to a lesser extent, cement-filled veinlets, micro-stylolites and voids. At thin section scale, dedolomitized samples mainly exhibit planar-euhedral, planar-subhedral and planar-porphyrotopic textures (calcified rhombs: 45 to 85 lm in diameter; Fig. 3C) and subordinately microcrystalline non-planar anhedral or macrocrystalline void-filling irregular textures. Pixel counting applied to microphotographs (Coimbra & Ol oriz, 2012a) revealed dedolomite contents ranging from 5% to 95% (Table SI-3). Drusy calcite filling voids and fractures is a common feature of the lower two-third of the section (0 to 49 m), whereas the upper part (53 to 74 m) is characterized by whitish, mud-dominated limestones with a chalky (microporous) appearance.

Bulk carbonate carbon and oxygen isotope stratigraphy
The bulk carbonate carbon isotope (d 13 C bulk ) record of the Jabal Madar section (Fig. 4) is characterized by numerous alternating negative and positive changes, with d 13 C bulk values ranging between À2Á8& and 3Á6& (mean: 1Á2&; SD: 1Á4&). In general, the lower part of the section (0 to 37Á4 m) shows stronger oscillations and overall lowered d 13 C bulk values (mean: 0Á4&; SD: 1Á3&)    and main fossil constituents and carbon and oxygen isotope stratigraphy of the Jabal Madar section, complemented with data from Pittet et al. (2002) and Sattler et al. (2005). Sequence stratigraphic interpretation after Pittet et al. (2002). At first glance, the low covariance between d 13 C and d 18 O values (r 2 = 0Á12) of all analysed bulk carbonate samples (0 to 74 m) points to the absence of major diagenetic alteration of isotope ratios. Splitting the data into two stratigraphic groups (A/B), however, provides r 2 -values that are indicative of a moderate (0 to 48Á8 m: r 2 = 0Á41) to strong correlation (49Á6 to 74 m: r 2 = 0Á65) of carbon and oxygen isotope values. when compared to the upper part (38Á9 to 74 m; mean: 2Á2&; SD: 0Á6&). The mean intra-sample d 13 C bulk variability is in the order of 0Á4& (SD: 0Á6&).
Considering 3-point running mean values, deposits of the lower part of the section (0 to 11 m) record a positive d 13 C bulk excursion with an amplitude of about 2Á4&, which is followed by a prolonged second positive 2Á1& carbon isotope excursion (11 to 30Á3 m). The latter excursion is terminated by a negative d 13 C bulk spike representing the lowest values of the carbon isotope record (À2Á8&). Background d 13 C bulk values are in the order of about À0Á5&. Upsection, the carbon isotope curve exhibits a prominent stepwise change (30Á3 to 37Á4 m) towards a maximum carbon isotope value of 3Á6& (37Á4 m). The subsequent chemostratigraphic segment (37Á4 to 74 m) is characterized by sinusoidal d 13 C bulk changes (amplitude: 1Á5&), which oscillate around a background value of about 2Á2&.
Bulk carbonate oxygen isotope (d 18 O bulk ) values range between À11Á1& and À1Á0& (mean: À6Á5.&; SD: 1Á9&). The mean intra-sample d 18 O bulk variability is in the order of 0Á6& (SD: 0Á7&). Strong d 18 O bulk oscillations characterize the lower part of the oxygen isotope curve (0 to 30Á3 m; mean À6Á9&; SD: 1Á9&), superimposed by a moderate trend to lower values (30Á3 m: À11Á1&). Upsection, the oxygen isotope curve exhibits a prominent positive trend (30Á3 to 31Á8 m) that reaches a first plateau (31Á8 to 48Á4 m; mean À4Á1&; SD: 1&), and finally a negative change (48Á4 to 53Á7 m) reaching a second plateau (53Á7 to 74 m; mean À7Á6&; SD: 0Á4&). The latter plateau is characterized by low-amplitude sinusoidal changes in d 18 O bulk . Cross-plots of bulk carbonate carbon and oxygen isotope data (Fig. 4) reveal a low covariance if all data are considered (0 to 74 m: r 2 = 0Á16). Based on petrographic (change towards mud-dominated 'chalky' limestones) and stable isotope features (lowering of d 18 O bulk background values), as observed in the upper part of the Jabal Madar section, two different stratigraphic groups of bulk carbonate samples (A/B) might be distinguished. Splitting the stable isotope data into these groups provides r 2 -values that are indicative of a moderate (A: 0 to 48Á8 m: r 2 = 0Á41) to strong correlation (B: 49Á6 to 74 m: r 2 = 0Á65) of carbon and oxygen isotope values.

Component-specific carbon and oxygen isotope variability
Carbon and oxygen isotope results of component-specific carbonate phases (matrix micrite, dedolomite, sparry cement, bivalve shell material) of selected samples derived from the Jabal Madar section are plotted in Figs 5 and 6.
Sampled cement phases are represented by drusy calcitic sparite that either fills bioclastic voids and intergranular pore space (spA) or fractures and voids (spB). In general, the distinguished cement generations bear clearly different isotopic signatures. SpB samples provide the most positive d 13 C (2Á6& to 4Á0&, mean: 3Á3&) and d 18 O values (À3Á9& to À2Á2&, mean: À3Á2&) of all sampled carbonate phases. In contrast, spA samples record a significantly wider range of both carbon (mean: À0Á1&, SD: 2Á3&) and oxygen isotope values (mean: À9Á6&, SD: 1Á7&). A cross-plot of d 13 C and d 18 O values (Fig. 6) allows two different groups of spA cements (spA-1, spA-2) to be distinguished. While spA-1 cements are moderately to strongly depleted in 13 C (À5Á7& to À0Á4&, mean: À2Á7&) and 18 O (À7Á9& to À14Á1&, mean: À10Á8&), spA-2 cements provide positive d 13 C values (0Á4& to 2Á4&, mean: 1Á2&) along with less variable d 18 O values (À9Á9& to À7Á5&, mean: 8Á9&).
Due to the often small size of cement-filled pore space, carbon (mean: 1Á8&, SD: 0Á7&) and oxygen isotope results (mean: À5Á8&, SD: 0Á9&) of a relatively large number (n = 25) of cement subsamples were discarded from further interpretations as they presumably represent a mixture of isotopic end-member compositions of spA and spB (Figs 5 and 6, see also Table SI-2), as will be further discussed.
Cathode luminescence characteristics, elemental geochemistry and strontium isotope analysis The objective of CL microscopy ( Fig. 7A-F) combined with elemental geochemistry and strontium isotope analysis ( Fig. 7G) is to provide independent evidence for the stable isotope-based interpretation of diagenetic realms, in which certain carbonate phases were formed. If combined with a (semi-)quantitative assessment of identified diagenetic products (i.e. dedolomite, sparite cements) at thin section scale, this approach allows the impact of the latter on the stable isotope signature of mud-supported and grain-supported bulk carbonate material to be evaluated. Component-specific elemental concentrations (Mg, Sr, Fe, Mn: normalized to a calcium content of 39Á7% for stoichiometric calcite), element/calcium and strontium isotope ratios are shown in Table 2. Only the most significant differences and similarities of analysed carbonate phases are reported here.
Cathode luminescence microscopy of matrix micrite reveals a commonly dull to moderately bright luminescence (Fig. 7A, B, and E). Dedolomite, often in the form of dedolomite rhombs, has a bright orange luminescence, contrasting with the duller surrounding matrix (Fig. 7C). SpA cements luminesce moderately to very bright orange to yellow and are characterized by a strongly zonal pattern ( Fig. 7A and B). SpB cements mainly inhabit fractures that clearly truncate grains and previous spA cements (Fig. 7D) and reveal a luminescence ranging from predominantly dull to moderately bright orange. In contrast to spA, spB cements are far less subjected to zoning. Both spA and spB cements frequently display partial alteration, expressed under CL by subordinate contributions of non-luminescent calcite and neomorphic microspar ( Fig. 7B and E) or by intra-crystalline micropores producing a spotted bright orange luminescence pattern ( Fig. 7A and D). Depending on the degree of recrystallization, bivalve fragments show a dull to dark luminescence (Fig. 7F).

Reliability of Jabal Madar bulk carbonate carbon and oxygen isotope data
At first glance, the Jabal Madar section exhibits a rather well-expressed bulk carbonate carbon isotope pattern (Fig. 4) that seems to allow for chemostratigraphic correlation purposes. Moreover, the low covariance between d 13 C and d 18 O values (r 2 = 0Á16) of all bulk carbonate samples points to the absence of major diagenetic alteration of isotope ratios. This prediction is obviously too simplistic, as the d 13 C bulk and d 18 O bulk cross-plot displays two different groups, each showing covariation between these proxies (Fig. 3). These groups comprise samples derived from the lower mixed mud-supported to grainsupported portion (0 to 48Á8 m: group A, r 2 = 0Á41) and the whitish-weathering predominantly mud-supported upper portion (49Á6 to 74 m: group B, r 2 = 0Á65) of the studied section. The observed difference of groups A and B is most likely related to primary contrasts in depositional setting (restricted vs. open lagoon; Pittet et al., 2002) resulting in a variable sedimentary contribution from marine and terrestrial sources, which ultimately promote textural differences such as the size, amount and distribution of initially available pore space (among others). These in turn influence hydraulic conductivity, accounting for differential susceptibility to early and/or later diagenetic processes. As for the covariant trends depicted for bulk samples (groups A and B), two lines of reasoning can be brought forward: (i) the geochemical record was severely affected by mixing-zone diagenesis and obtained results no longer reflect original palaeoenvironmental conditions; (ii) sampling different generations of carbonate materials (matrix micrite, cements, dedolomite) may also produce such covariation, by which the original signal is still recorded at least in some carbonate phases (Allan & Matthews, 1982). In either case, differential diagenetic pathways must be explored in order to provide a clear understanding of the obtained bulk carbon and oxygen isotope patterns and determine if further interpretations based on the stratigraphic evolution of d 13 C BULK should be pursued. In particular, the occurrence of dedolomite and sparite cement phases requires an indepth evaluation as to their relative contribution to the bulk carbonate stable isotope archive. By means of a component-specific petrographic, elemental and stable isotope analysis, quantitative arguments can be used to extract an objectively reliable near-primary (shallow marine) carbon isotope signal which will allow a refined integrated stratigraphic framework to be established.

Component-specific geochemical and petrographic properties
It is widely accepted and demonstrated that d 13 C records are generally less affected by diagenetic overprint than the more sensitive d 18 O signal, as carbon isotopes are less vulnerable to changes in the water-rock ratio and suffer only minor alteration towards lower d 13 C values with increasing burial depth and associated increasing temperatures (Emrich et al., 1970;Marshall, 1992). Diagenesis is thus often conservative when carbon isotopes are concerned, from sustaining absolute original values, to at least preserving the original trends along a studied time frame (Frank & Lohmann, 1996). In both cases, d 13 C records reflect original fluctuations in response to palaeoenvironmental conditions. Here, well preserved bulk carbon isotope signals were identified by focusing on the geochemical and petrographic properties of component-specific diagenetic products and eliminating their potential contribution to the obtained bulk d 13 C curve.
Both a stratigraphic and a cross-plot of carbon and oxygen isotope values illustrate the high variability in component-specific stable isotope signatures (Fig. 6). Although the elemental and isotopic composition of biogenic hard parts (Immenhauser et al., 2008;Sch€ one, 2008) are known to severely depend on metabolic processes, previous studies have shown that bivalve shells might act as a suitable archive for mid-Cretaceous secular carbon isotope changes (Huck et al., 2012;Huck & Heimhofer, 2015). Unfortunately, the petrographic inspection of bivalve shell fragments derived from the Jabal Madar section provides clear evidence for a moderate recrystallization and partial silicification of the fibrous low-Mg calcite ultrastructure. The majority of bivalve and matrix micrite carbon isotope values, in contrast to the highly variable bulk carbonate values, are close to the expected Barremian mean open marine d 13 C value of about 1 to 3& as recorded in Tethyan and Pacific pelagic sections (Weissert et al., 1985;Jenkyns, 1995;Weissert & Erba, 2004;Godet et al., 2006;Sprovieri et al., 2006;Prokoph et al., 2008;Bodin et al., 2009;Wohlwend et al., 2017). At first glance, this would imply that the most fine-grained portions of the Jabal Madar section escaped considerable overprint of their d 13 C signal by diagenetic fluids and thus, might be regarded as a suitable archive for secular carbon isotope trends of dissolved inorganic carbon (DIC). As stated by Immenhauser et al. (2008), however, neritic ooze 'cannot be considered to be free of diagenetic artefacts', as this polygenetic and therefore partly metastable material is expected to be prone to at least syn-depositional (pore-water) diagenesis (Patterson & Walter, 1994;Sanders, 2003;Walter et al., 2007;Coimbra & Ol oriz, 2012b). But such effects may result in only subtle changes to the original geochemical signals, thus even stabilized carbonates can provide reliable palaeoenvironmental records (Coimbra et al., 2009;Vincent et al.,  ratios of all analysed subsamples, in contrast to a cluster of (arguably) meteorically altered matrix micrite providing considerably higher Fe/Ca ratios. A clear covariation in Sr/Ca ratios and strontium isotope ratios is observed among selected matrix micrite samples, an observation that points to the influence of radiogenic strontium related to siliciclastic weathering.

2017
). This possibility was tested by comparing carbon and oxygen isotope values, elemental contents and CL characteristics of matrix micrite (sensu stricto) with those of all other carbonate components identified in samples from the Jabal Madar section (Fig. 6), aiming to extract a highly detailed diagenetic evolution of the carbonate materials under study. Matrix micrite, in most of the thin sections, appears dull to moderately bright under CL. This often documented luminescence pattern mirrors the diagenetic stabilization of matrix micrite soon after deposition, either close to the sediment/water interface (Coimbra et al., 2009) or in the shallow burial diagenetic domain. This interpretation is in line with the majority of matrix micrite d 13 C and d 18 O signatures, which are in the range of penecontemporaneous sea water isotope values (Coimbra et al., 2009) or only slightly deviated towards (more negative) burial diagenetic signatures (Christ et al., 2012).

Processes affecting carbon isotope signals
In line with sedimentological evidence for a restriction from the open marine realm and transient episodes of subaerial exposure (root traces, laminated mudstones with desiccation cracks or single-species (ostracod) assemblages), the stable isotope pattern ('inverted J") of a small cluster of bulk and matrix micrite samples (n = 8) documents the stratigraphically limited influence of mild meteoric diagenesis (Allan & Matthews, 1982;Meyers & Lohmann, 1985;Lohmann, 1988;Figs 5 and 6). The lack of a distinct and more profound exposure-related stable isotope pattern (strong decrease in d 13 C values related to meteoric fluids bearing 12 C-enriched soil-gas CO 2 ) might be explained by (i) the initial palaeogeographic and climate setting, (ii) transient exposure periods inhibiting the formation of thick soil horizons, (iii) the erosion of the meteorically altered sediment pile during subsequent transgression and/or (iv) post-depositional dissolutionprecipitation processes that might have masked the meteoric precursor signals (cf. Coimbra et al., 2016;Godet et al., 2016). Subsamples with abundant microcrystalline replacive dedolomite (≥75%) as well as a cluster of sparite A subsamples (spA-1) provide the most negative carbon (and oxygen) isotope values. The impact of the latter diagenetic products on the bulk stable isotope composition is clearly related to their abundance, since samples with the lowest dedolomite contents (5 to 10%) record carbon (and oxygen) isotope values overlapping the isotope range of the best-preserved matrix micrite samples. This observation clearly reflects the mixing of the isotope signatures of the dedolomite end-member and the host rock, accounting for the fair correlation of both dedolomite subsamples and bulk samples of group A (Fig. 4; see also Table SI-3). Moreover, the observed (recent?) selective dissolution of replacive very small dedolomite crystals, although less abundant, appears to be associated with cyclic positive carbon isotope changes and thus results in rather strong correlation of isotope signatures of 'chalky' samples representing group B (Fig. 4). As carbon isotope values of dedolomite-rich samples are considerably depleted (down to À5Á5&), involved burial fluids might have been sourced by 12 C derived from the thermal breakdown of buried organic matter that escaped decomposition in the predominantly fine-grained portions of the Jabal Madar section. The alternative of early diagenetic meteoric fluids charged with soil-zone CO 2 that might have acted as donor of isotopically light carbon seems at first glance less likely. This alternative would imply the development of thick soil horizons and related mature karst features during a time (Late Barremian) of predominantly arid climate (Ruffell & Batten, 1990;Godet et al., 2008;F€ ollmi, 2012;Amodio & Weissert, 2017). But even if erosion of former soil horizons occurred, the predicted oxygen isotope composition of meteoric waters in low latitudinal coastal settings contradicts this possibility (as further discussed).
Sparite A (spA) cements reveal two different carbon isotope signatures, from strongly to only slightly lowered values when compared to the range of presumed well-preserved marine signals. This feature reveals the influence of different sources of fault fluids, which may relate to multi-stage faulting events or merely to different directions of faulting (bed normal, parallel or oblique veins) promoting the circulation of different fluids (Agosta et al., 2008).
Sparite B (spB) samples provide enigmatic, considerably higher carbon isotope signatures (up to 4&) when compared to any other carbonate material and to the expected Barremian (shallow-) marine signal (Fig. 6). The timing of late veining and concomitant infill of available pore space of spB cements is hard to establish, but major spB veins cross-cut most of the remaining diagenetic features, so their origin is probably related to fluids circulating during or after sub-recent exhumation (and creation of secondary pore space due to dissolution processes) of the Cretaceous succession during the Miocene (Fig. 8). Due to similar carbon and oxygen isotope compositions of all spB cements and their encasing matrix micrite (not influenced by meteoric alteration), the influence of rockbuffered fluids is envisaged, but the clear increase in carbon isotope values merits attention (Fig. 5). The persistently higher carbon isotope signature is not abrupt enough to consider processes related to fermentation of organic acids, CO 2 reduction or excessive evaporation (Clayton, 1994). Alternatively, a slight carbon isotope fractionation is here proposed, owing to lowered pCO 2 in fluids circulating at shallow depth, which is reasonable during pressure release phenomena (Shemesh et al., 1992;Hassan, 2011). Several reports of slightly heavier d 13 C values along late diagenetic calcite and even dolomite veins show a similar pattern (Morad et al., 2010;Vandeginste et al., 2013;Arndt et al., 2014;Balsamo et al., 2016). In any case, these cements are mostly well-expressed with respect to their textural (predominantly cross-cutting thick veins), CL (dull) and elemental characteristics (low Fe and Mn contents) (Figs 7 and 8). Bulk samples, in particular if mud-supported, are thus unlikely to be influenced by the geochemical signal of spB. As mentioned earlier, however, CL indicates that a limited number of cement-rich, grain-supported samples host interfingering sparite cements of spA and spB and/or provide evidence for sparite cement diminution by neomorphic microsparite patches (Fig. 7F). Carbon isotope values of the corresponding cements indeed reflect a mixture of spA and spB end-member compositions (Fig. 5). Consequently, the variable contribution of spB cements and microspar might shift the bulk carbon isotope signature towards higher values.

Processes affecting oxygen isotope signals
Regarding oxygen isotope values, the least depleted carbonate materials are matrix micrite, spB cements, selected bivalve shells and very weakly dedolomitized samples (Fig. 6). These partially overlap the expected marine signal, tailing towards the uppermost range of spA-2 cements. Selected bivalve shells showing partly preserved growth increments display rather low d 18 O values (À7Á0 to À4Á5&) and strontium concentrations (336 to 708 p.p.m.), both geochemical signatures pointing to the alteration of shells in the shallow burial realm by marine fluids, despite their rather low iron (mean: 12 p.p.m.) and manganese (mean: 20 p.p.m.) concentrations. In line with considerably lowered d 18 O values of spA-2 cements, this confirms the progressive influence of fluids with elevated temperature in a rock-buffered system (Moore, 1985;Choquette & James, 1987;Allan & Wiggins, 1993;van der Kooij et al., 2009), generating more depleted d 18 O values in carbonate phases precipitating from marine fluids in the intermediate burial realm.
Notably, a specific group of sparite cements (spA-1) and carbonates with a high abundance of dedolomite are strongly depleted in oxygen isotope values (down to À14&), along with also very negative d 13 C values (down to À6&). Perhaps the best explanation for the observed strongly depleted carbon and oxygen isotope values, as opposed to the thermal breakdown of buried organic matter in the deep burial realm, is the burial and subsequent expulsion of soil-zone CO 2 -enriched meteoric waters (de Caritat & Baker, 1992;Moss & Tucker, 1995;Hendry, 2002;Frazer et al., 2014). A precursor meteoric d 18 O signature of À5& would then translate into a maximal burial depth of 1Á9 km, well within the range of a previously estimated intermediate burial depth of 1 to 4 km for Cretaceous deposits exposed along the Adam foothills at Jabal Madar and Jabal Qusaybah (Hanna, 1990;Mozafari et al., 2015). This alternative is also favoured against the contribution of thick former soil horizons, because the recorded shift in oxygen isotope ratios of meteoric waters towards values of up to À14& is incompatible with the minimal effect of isotopic fractionation during evaporationprecipitation (Rayleigh Effect) at low latitudinal coastal settings (Dansgaard, 1964;Anderson & Arthur, 1983;Lohmann, 1988).

Establishing isotopic threshold signals
Although the contribution of burial-related diagenetic processes evidently influenced the oxygen isotope signature of Jebel Madar bulk carbonate, they are generally not believed to have a major impact on carbon isotope geochemistry (Banner & Hanson, 1990;Veizer et al., 1999), in particular if cements are volumetrically of low importance. In fact, even bulk carbon isotope values recorded by cement-rich (spA-2) grain-supported samples largely resemble those of matrix micrite samples (1 to 3&) and, more importantly, of selected best-preserved bivalve shells. However, the evaluation of component-specific isotope signatures (bulk vs. spA-2) revealed that enhanced spA-2 contents in the order of about 25% (sample JM40.5) result in a lowering of the d 13 C bulk value by about 0Á7& (Fig. 7D; Table SI-2). At limited stratigraphic intervals, this lowering of d 13 C bulk values is counteracted by the partial replacement of spA-2 cements by a later diagenetic cement (spB) bearing considerably higher carbon (and oxygen) isotope values.
In order to provide solid evidence for the usefulness of bulk carbon isotope values and overall stratigraphic trends for further chemostratigraphic and palaeoenvironmental interpretations, critical thresholds established by the indepth analysis of component-specific samples have to be applied. In particular, the stable isotope signatures of mudsupported, often dedolomite-rich (≥75%), and cement-rich grain-supported portions (group A) of the Jabal Madar section merit attention. These samples either experienced mild early meteoric diagenesis (10 to 12 m, 28 to 36 m; Fig. 5) or severe diagenetic alteration by buried meteoric waters during the mesogenetic stage (Fig. 5). It is therefore concluded that the variable contribution of dedolomite and sparite A largely dictates the diagenetic trends obtained for bulk carbonate samples. In order to confidently establish a carbon isotope threshold that is not affected by either of these cement phases, bulk stable isotope signatures overlapping with those provided by meteorically altered samples, strongly dedolomitized samples and spA-1 samples (d 18 O <À7Á8&, d 13 C <0Á9&) are discarded from further stratigraphic and palaeoenvironmental interpretations. Moreover, the possible influence of spB cements on the bulk carbonate record is lowered by excluding samples with d 13 C values above 3Á1&. Notably, the filtered bulk carbonate samples (107 out of 202 samples) fall within matrix micrite sensu stricto (not influenced by meteoric fluids) and selected bivalve shell isotope signals.
In summary, the in-depth component-specific petrographic and geochemical approach presented here allows the Jabal Madar shallow-water bulk carbonate carbon isotope archive to be critically assessed. The resulting 'cleaned' carbon isotope record is expected to have preserved a near-primary pattern of superimposed global d 13 C trends and inflexion points.

Integrated Barremian-Aptian shallow-water stratigraphy of Northern Oman
The Barremian-Lower Aptian portion of the Jabal Madar section shows a characteristic sedimentary stacking pattern, which is easily recognizable in the field as repeated alternation of (i) predominantly thin bedded partly argillaceous (upper Lower Kharaib and Hawar members) and (ii) more massive and partly cross-bedded intervals (lower Lower Kharaib, Upper Kharaib and Lower Shu'aiba; Fig. 3A). The latter depositional cycles are welldocumented throughout the Arabian platform both in outcrops and in the subsurface and referred to as thirdorder sequences I (AP Bar1), II (AP Bar2) and III (AP Apt1-4) (Harris et al., 1984;Hughes-Clarke, 1988;Sharland et al., 2001;Pittet et al., 2002;van Buchem et al., 2002;Strohmenger et al., 2006;van Buchem et al., 2010). Following a revised Barremian-Aptian orbitolinid biostratigraphic zonation of the eastern Arabian Plate (Schr€ oder et al., 2010), the occurrence of the short range index fossils M. arabica (see also Simmons, 1994) and E. transiens in the transgressive deposits of the upper Lower Kharaib Member at Jabal Madar (sequence II) is indicative of an early to middle Late Barremian age, whereas the mass occurrence of P. lenticularis in the early transgressive Hawar Member points to an early Early Aptian age.
The 'filtered' bulk carbonate carbon isotope stratigraphic record established here (complemented with data from Sattler et al., 2005) allows for a chemostratigraphic characterization of the combined biostratigraphicsequence stratigraphic scheme Schr€ oder et al., 2010). Initially, carbon isotope stratigraphy is applied to correlate sequences I to III on a regional (Oman) scale. Unfortunately, continuous Barremian-Aptian shallow-water carbon isotope records are scarce throughout the Arabian platform. Two exceptions are the uppermost Barremian-Lower Aptian Wadi Mu'aydin d 13 C record from the southern rim of the Jebel Akhdar dome in northern Oman (Wilson, 1969;Glennie et al., 1974;Simmons & Hart, 1987;Simmons, 1990;van Buchem et al., 2002) and the Upper Barremian-Lower Aptian Huqf d 13 C record (section S018) from south-eastern Oman (Immenhauser et al., 2004;Sattler et al., 2005).
In the Huqf area ( Fig. 9), partly dolomitized tidal flat deposits representing the highstand of sequence I display lowered and strongly fluctuating carbon isotope values. This pattern has been associated with local water mass 'ageing' and superimposed repeated subaerial exposure events in the intertidal realm (Sattler et al., 2005). In the relatively more distal Jabal Madar section, a limited number of matrix micrite subsamples within transgressive and early highstand deposits of sequence I also record meteorically induced negative carbon isotope changes in the order of up to 3&. The arguable meteoric origin of these negative d 13 C excursions is supported by enhanced iron (mean: 641 p.p.m.) contents recorded by micrite samples and the occurrence of delicate spar-filled root casts and thin laminated mudstone layers (Table 2; Figs 7 and 9).
The overlying largely grain-supported high-energy deposits representing the late highstand of sequence I, in contrast, lack clear isotopic and petrographic evidence for a substantial overprint by meteoric fluids. Considering CL characteristics (showing dominance of spA-2 cements), this might be explained by burial-related dissolution-precipitation processes.
Chemostratigraphically, sequence II both at Jabal Madar and Huqf is represented by cyclic carbon isotope fluctuations superimposed on a broad positive d 13 C bulge. At all considered localities including Wadi Mu'aydin, sequence II is terminated by a prominent negative incursion immediately beneath the onset of the overlying Palorbitolina lenticularis-rich Hawar Member. At Jabal Madar and Huqf, this change is associated with laterally extensive composite surfaces (CS) that bear evidence for both a subaerial exposure and a marine hardground stage (Sattler et al., 2005). This observation supports the notion of a major sea-level fall and subsequent long-lasting exposure (10 5 years) of the Arabian carbonate platform at the Barremian-Aptian transition Al-Husseini & Matthews, 2010).
At Jabal Madar and Wadi Mu'aydin, the transgressive deposits of sequence III record gradually increasing carbon isotope values (Hawar Member), which are again capped by an abrupt negative 2& change in d 13 C at the transition towards the Lithocodium-Bacinella bearing deposits of the Lower Shu'aiba Member. Following previous work in Oman Sattler et al., 2005;Huck et al., 2010), the observed prominent negative d 13 C spike represents the chemostratigraphic segment C3 sensu Menegatti et al. (1998), which precedes oceanic anoxic event (OAE) 1a. The Huqf section provides a rather similar pattern, but due to the restricted intertidal conditions, d 13 C values of the Hawar Member are strongly overprinted by meteoric carbon isotope signatures (Sattler et al., 2005). In contrast to the rather well-established, integrated stratigraphic framework for the Upper Barremian-Lower Aptian carbonate platform deposits exposed at Jabal Madar, Wadi Mu'aydin and the Huqf area, the long range of the Hauterivian-Early Barremian Permocalculus inopinatus biozone (Simmons, 1994) hampers a precise carbon isotope-based age assignment for the Upper Lekhwair Formation and Lower Kharaib Member at Jebel Madar, in particular, as additional carbon isotope stratigraphic results from Wadi Mu'aydin and the Huqf area are presently not available.
Since the petrographic and geochemical properties of certain carbonate phases clearly indicate a rather complex multi-stage diagenetic alteration of the considered interval at Jabal Madar (Fig. 8), the aim is to establish a best-fit chemostratigraphic framework by considering Tethyanwide neritic and pelagic major carbon isotope and biosedimentation changes (Fig. 10). The Northern Tethyan Cluses section serves as a chemostratigraphic shallowwater reference, as the Barremian-Aptian carbon and strontium isotope pattern recorded at this locality allowed precise correlation with well-dated pelagic sections (Angles, Gorgo a Cerbara) in the Vocontian and Umbria Marche basins (Busnardo, 1965;Godet et al., 2006;Sprovieri et al., 2006;Bodin et al., 2009;Huck et al., 2011Huck et al., , 2013Stein et al., 2011). An age-calibrated carbon and strontium isotope reference frame (Huck & Heimhofer, 2015) builds on this platform-to-basin correlation, providing a well-constrained Barremian pattern with distinct chemostratigraphic tie points including (i) a positive change in pelagic d 13 C background values known as the Mid-Barremian Event (MBE), (ii) a Late Barremian gradual (pelagic) positive carbon isotope bulge, and (iii) the above-mentioned negative variation in d 13 C at the Barremian-Aptian boundary.
The MBE has been arguably associated with a relative increase in the basinward export of platform-derived aragonitic detritus     Table 2). Please refer to the legend of Fig. 4 for information on symbols used in the rock column.
2006) in the Northern Tethyan realm, although sedimentological evidence seems to promote enhanced black-shale formation in the Boreal (Lower Saxony Basin: Malko c & Mutterlose, 2010) and central Tethyan realms (Umbria Marche Basin: Sprovieri et al., 2006) as a potential driver of this change in pelagic carbon isotope values. In shallow-water sections, the MBE is generally less well-constrained due to the overall large variability in d 13 C, which is in particular related to the exposure-related meteoric overprint of deposits in the prelude and aftermath of the MBE (Di Lucia et al., 2012;Huck et al., 2013). At all considered reference sections, however, the chemostratigraphic pattern beneath the MBE-positive values is characterized by a two-fold negative d 13 C excursion, the latter linked to a major subaerial exposure event in the Northern Tethyan realm (Huck et al., 2013). Notably, a similar pattern of transient (meteorically induced) negative values and a subsequent prominent positive excursion in d 13 C is recorded at Jabal Madar (30 to 38 m; Fig. 9).
The onset of the gradual positive carbon isotope bulge in the Umbria Marche Basin coincides with the onset of rhythmic black-shale deposition (Sprovieri et al., 2006). At Cluses, this positive trend appears to be attenuated by local carbon cycling processes that shift the carbon isotope record towards more negative values. There, a return to more open marine conditions is documented by a rapid and prominent d 13 C shift within transgressive Palorbitolina lenticularis-rich deposits ascribed to the Hemihoplites feraudianus ammonite zone. Due to the influence of platform-derived carbonate detritus, the adjacent hemipelagic Angles section in the Vocontian provides a similar carbon isotope pattern (F€ ollmi et al., 2006). The stepwise positive carbon isotope trend as displayed by the shallow-water Jabal Madar record shares similarities with the Umbria Marche d 13 C pattern. In contrast to the Cluses section, slightly argillaceous Palorbitolina-rich deposits occur at Jabal Madar already at the onset of the positive carbon isotope bulge (~Heinzia Sayni ammonite zone). An oyster fragment derived from this interval at Jabal Madar provided a strontium isotope value (0Á707556) that slightly deviates from the expected range of 87  ?  δ 13 C (‰) 0 1 2 3 II Fig. 10. Carbon isotope based correlation of the 'filtered' Jabal Madar record with age-calibrated neritic (Cluses) and basinal (Angles, Gorgo a Cerbara) reference records Sprovieri et al., 2006;Huck et al., 2011;Stein et al., 2011;Huck & Heimhofer, 2015). Note extremely low Barremian carbonate preservation rates as calculated from the Jabal Madar section. Thick grey lines represent 3-point moving average curves. Dashed red lines highlight a positive shift in pelagic (and arguably neritic) d 13 C background levels at the onset of the Late Barremian. The stratigraphic position of major discontinuity surfaces (CS) and orbitolinid-rich levels are indicated. 2011). Given the observed increased shedding of clay in the upper Lower Kharaib Member at Jabal Madar, as well as the clear radiogenic strontium isotope signature as recorded by matrix micrite samples (0Á707923 to 0Á708947, Fig. 7), the observed shift towards more positive values is interpreted to reflect the influence of riverine input of radiogenic strontium related to continental silicate weathering. In accordance with previous SIS studies based on oyster shells in Cretaceous coastal settings (Burla et al., 2009;Heimhofer et al., 2012;Horikx et al., 2014), a moderate radiogenic impact on the 87 Sr/ 86 Sr signature (ca 5 9 10 À6 ) of the best-preserved shell selected here is assumed. The measured value might therefore support an early Late Barremian age. Without additional strontium isotope samples that might record secular trends of radiogenically influenced shallow marine 87 Sr/ 86 Sr, however, this age assignment remains tentative.
In summary, the integration of sequence stratigraphic and biostratigraphic data (Pittet et al., 2002 and this study) with the evaluated bulk carbonate d 13 C record enables a chemostratigraphic characterization of the Jabal Madar section. Considering this integrated stratigraphic framework, the diachronous nature of Late Barremian orbitolinid mass occurrences as recorded by northern and southern Tethyan transgressive carbonate platform deposits becomes evident. However, a precise chemostratigraphic age assignment of Hauterivian to Lower Barremian shallow-water carbonates of the Arabian carbonate platform is still problematic due to (i) the low-resolution of shallow-water biostratigraphic zonation schemes and (ii) the lack of carbon isotope records along the Arabian carbonate platform, which hamper distinguishing between local (diagenetic) and global carbon isotope perturbations and the identification of carbon isotope gradients across the platform. Moreover, extremely reduced Barremian carbonate preservation rates (ca 0Á012 mm y À1 ) reveal that large parts of the carbon isotope signal at the Jabal Madar locality is lost in discontinuity surfaces. The platformwide assessment of component-specific carbon isotope records, however, has the potential to solve these stratigraphic uncertainties, in particular if combined with SIS, which focuses on secular rather than short-term changes in marine 87 Sr/ 86 Sr.

CONCLUSIONS
Bulk carbonate stable isotope signatures recorded by a tropical mid-Cretaceous (Barremian-Aptian) shallowwater limestone succession (Jabal Madar section, Oman) provide evidence for a differential multi-stage diagenetic alteration. This is revealed by two stratigraphic clusters of samples providing significant correlations of carbon and oxygen isotopes. An in-depth petrographic (CL microscopy) and geochemical evaluation (C, O, Sr isotopes, trace elements) of different carbonate phases shows evidence for a stratigraphically variable and often facies-related impact of different diagenetic fluids on the bulk-rock stable isotope signature. The presence of abundant replacive dedolomite in mud-supported limestone samples forced negative carbon and oxygen isotope changes that are either associated with the thermal breakdown of organic matter in the deep burial realm or the expulsion of buried meteoric water in the intermediate burial realm. Sparite cements filling intergranular pores, bioclastic voids and fractures (spA) evidence intermediate to (arguably) deep burial diagenetic conditions during their formation, owing to different timing or differential faulting promoting the circulation of fluids from different sources. In contrast, a second group of sub-vertical vein-filling sparite cements (spB) reveal a rock-buffered diagenetic fluid composition with an intriguing slight enrichment in carbon isotope values, probably due to fractionation during pressure release in the context of the Miocene exhumation of the carbonate platform deposits under study. The superposition of stable isotope signatures of identified carbonate phases causes a complex and often noisy bulk carbon isotope pattern that offers multiple chemostratigraphic interpretations due to the low-resolution of biostratigraphic constraints. The componentspecific isotopic and petrographic approach defined threshold values for d 13 C (>0Á9&, <3Á2&) and d 18 O (>À7Á8&) that allowed the diagenetically least altered samples to be extracted and thus identify the most reliable 'primary' bulk carbon isotope signatures. The integration of the 'filtered' carbon isotope curve with lithostratigraphic, sequence stratigraphic and biostratigraphic data allows the Jabal Madar section to be compared with two (regional) Upper Barremian-Lower Aptian shallow-water sections in Northern (Wadi Mu'aydin) and southern Oman (Huqf). Although extremely low carbonate preservation rates at all considered localities reveal that most of the carbon isotope signal is lost in discontinuity surfaces, characteristic long-term trends as observed in the Jabal Madar record allow for a (tentative) chemostratigraphic correlation with stratigraphically well-constrained Tethyan reference carbon isotope curves. Chemostratigraphic tie points include (i) a positive shift in pelagic and neritic background d 13 C values referred to as the Mid-Barremian Event, (ii) a long-lasting Late Barremian-positive carbon isotope bulge and finally (iii) two distinct negative d 13 C values marking major palaeoceanographic changes at the Barremian-Aptian boundary and at the onset of the Early Aptian Oceanic Anoxic Event 1a.

Supporting Information
Additional Supporting Information may be found online in the supporting information tab for this article: Table S1. Bulk carbonate carbon-and oxygen isotope results (Jabal Madar section). Table S2. Component-specific carbon-and oxygen isotope variability (Jabal Madar). Table S3. De-dolomite contents and stable isotope data of selected subsamples (Jabal Madar).